An Estimation of Distribution Algorithm Hybridized with Simulated Annealing for Makespan-Aware Task Scheduling in Fog Computing
Task scheduling in heterogeneous fog–cloud computing requires allocating independent tasks to virtual machines while reducing completion time. This article investigates a discrete hybrid of an estimation-of-distribution algorithm (EDA) and conventional simulated annealing (SA), denoted EDA–SA, for the single-objective makespan-minimization model. The SA refines selected solutions through relocation and swap neighborhoods with a fixed geometric cooling schedule. We evaluate the proposed approach on synthetic tasks uniformly distributed between 1,000 and 20,000 million instructions and on the original GoCJ trace, at five and ten task scales, respectively. We compare against different metaheuristics and five other algorithms in the literature. Makespan, degree of imbalancing, and convergence are recorded over 30 independent runs per task scale.
Authors
- Nabeel Ali
Institutions
- University of Kufa (IQ)
Publication Details
- Journal
- Iraqi Journal of Information and Technology Systems
- Published
- 2026-11-12
- DOI
- https://doi.org/10.112222/ijits.v1.i2.25985
- Primary Topic
- Cloud Computing and Resource Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00